1076 — Levels and Variation in Overuse of Colorectal Cancer Screening in VHA
Partin MR, Nelson DB, and Powell AA, Minneapolis VAMC; Burgess JF, Boston VAMC; Bangeter A, and Halek K, Minneapolis VAMC; Fisher DA, Durham VAMC;
As colorectal cancer screening (CRCS) rates have surpassed national goals in the Veterans Health Administration (VHA), attention has turned to potential overuse of CRCS. Because Fecal Occult Blood Testing (FOBT) is the CRCS method most commonly used in VHA, we examined levels and correlates of FOBT overuse in a nationally representative VHA sample.
The sample included 3,025 CRCS-eligible patients responding to a 2007 survey of CRCS behavior conducted in 24 VHA facilities. We combined patient demographics from this survey with VHA administrative data on CRCS procedures and outpatient visits received by respondents between 2003-2009 in order to estimate FOBT overuse, and examine variation by region, facility complexity, and patient age, race, gender, education, income, and outpatient utilization. We coded FOBT procedures as overused if they were conducted sooner than recommended by guidelines (<10 months after prior FOBT, <9.5 years after prior colonoscopy, <4.5 years after prior barium enema). We used multinomial logistic regression models to examine facility and patient variation in FOBT overuse by reason (sooner than recommended after prior FOBT; sooner than recommended after colonoscopy, barium enema, or a combination of procedures). All estimates adjusted for the clustering of procedures within patients and facilities.
Of the 4,950 FOBTs received by our cohort, 1,003 (20%) met our criteria for overuse. Of these, 357 (36%) were conducted sooner than recommended after FOBT, and 646 (64%) sooner than recommended after other procedures (colonoscopy, barium enema, or combination). FOBT overuse varied significantly by facility (8-32%, p <0.0001), but not patient demographic characteristics. The odds of overuse after prior FOBT varied by region (OR 0.29 in Midwest versus Southwest, p = 0.01), facility complexity (OR 1.81 for lowest versus highest complexity, p = 0.002), and increased with number of outpatient visits (OR 1.16, p <0.0001). The odds of overuse after other procedures varied by region (OR 0.41 p = 0.008 in Midwest; 0.43 p = 0.009 in Northwest, versus Southwest), and outpatient visits (OR 1.16, p <0.0001).
Many VHA FOBTs are performed sooner than recommended, and this overuse varies significantly by facility characteristics and health care utilization.
The significant facility variation in FOBT overuse suggests efforts to address this problem will likely require system-level solutions.